2024-05-01 Formalization Working Group Meeting notes

• Craig submitted 5 wolfcon sessions related to the PoC
  
  • Can share abstracts if folks are interested but they may get updated when the time is closer
• Working on overview that uses some of the slides from the session and has more detail. Some time next week or the week after will distribute. More concrete if not totally complete  documentation.
• Not much RFC progress
• One of the EPAM architects is looking at the app composition spreadsheets and will make some refinements/suggestions. Can share what is found when he's ready. Probably more of a technical review than product.
• Will bring questions from the last minutes that weren't answered over to the feedback page
• Could also add FAQs to the overview document- maybe best to do after the next meeting
• Can also cross reference the page to the minutes

• Anything folks want to bring forward from last week's meetings
• From TC
  • How the architectural change relates to the charge of application formalization
  • Any change in scope?
  • Things are kind of "all mixed up" ie the architectural stuff and the app formalization are kind of smushed together
  • How/if to transition - communication, facilitation lie with this group
  • Does  the community want all of this?
  • RFCs are documentation and decision records for understanding what happened and why
• Clarification on what the plan is for the Ramsons release
• Process for decision making on this scale
  • Previously decisions often made by work done and promoted
• Hard to answer what ifs/ how long transition period
• Shift in community toward acceptance with more information being available
• Risk doesn't seem architectural, the risk is that there is a lot of code that represents reimplementation of functionality that is already in Okapi and the auth modules
  • Risk of new code replacing code that has been stable for a couple of years. How to discuss this risk?
  • Introduction  of the suite of applications that manage module relationships and tenant relationships and integration of key cloak. Request routing and authorization happening in different ways, introduction of side cars. Module level not a lot of changes if any. Changes are at the FOLIO core/foundational level. New code, less proven than current but LC is going live on it, have done a lot of internal testing. Need to share some results and test cases to help build trust in the new core. How else to address?
  • Community should pay attention to the right risk. Risk isn't necessarily problematic, just point out where hosting providers see the risk.
• EBSCO going first means functionality will be tested by EBSCO quite thoroughly, high motivation for testing. EBSCO going first helps with everyone else's risk.
• Hosts also need to understand the new components, side cars, etc. Training on the tooling.
• All  of these things need to be compared with the concerns people have about the architectural branches
• some hurdles around CI they are still working on. Sysops/devops may have the biggest hurdles. hosting  providers are more unique than similar. won't be the same recipe for each provider, and that will be a challenge
• tried to make easy to adopt, but hosts will have work to adopt, they will have to get familiar with the architecture, new skills, deployment is environment specific
• most hosts unlikely to be ready for ramsons
• mechanism for deciding when/if, votes by each of the councils?

Draft summary document: https://docs.google.com/document/d/1rc4xgJQsNFlXoHW6lAI7MbF8MzaWDb9QW2eryiiyNTo/edit

Draft of answered questions still to come

Next steps

• Type your task here, using "@" to assign to a user and "//" to select a due date

Zoom chat:

00:01:33	Jenn Colt:	https://folio-org.atlassian.net/wiki/spaces/CC/pages/175277417/2024-05-01+Formalization+Working+Group+Meeting+notes
00:22:25	Marc Johnson:	We’ve never made such wholesale changes on this scale before (except for early on)
00:22:45	Marc Johnson:	RFCs as documentation is different to RFC as process
00:26:17	Wayne Schneider:	Indeed, no one probably can adopt this tomorrow, including the Library of Congress :-)
00:30:48	Craig McNally:	"tomorrow" was hyperbole.  Really I meant "in the sort term"
00:34:33	Marc Johnson:	IMO the control the community has is the if / when decisions
00:37:56	Marc Johnson:	Another aspect is familiarity with the architecture
00:38:11	Marc Johnson:	The if decision has to come before the when
00:44:43	Wayne Schneider:	LOL to be clear I have confidence in the EBSCO team
00:53:31	Wayne Schneider:	FOLIO operations is not terribly mature as a discipline
00:53:58	Marc Johnson:	What Craig is saying is what I meant by operations familiarity with the new architecture (and the new tools)
00:54:17	Tod Olson:	Reacted to "What Craig is saying..." with ✔️
00:59:38	Marc Johnson:	The way the RFCs are structured now are not well suited to adopt eureka or not decision
01:00:37	Craig McNally:	thanks guys, I have to drop for another meeting
01:05:50	Marc Johnson:	I should say I’m biased against recommendations